Congestive heart failure (CHF) is a debilitating condition that afflicts 4.8 million Americans with an increasing incidence. Despite efforts at preventative cardiac care, the incidence of CHF is increasing because the average age of survival in the population is increasing and because more people are surviving their first heart attack. Pharmacological therapy and electrical stimulation therapies are improving, yet many patients still reach end-stage heart failure. Heart transplant is available for some patients with the most severe heart failure, however the supply of donor hearts is not adequate to meet demand and there are many complications associated with immunosuppression. Aberrant growth and remodeling are evident in CHF, and it is likely that growth and remoding are primary contributors to infarct expansion, myocardial scaring, and ventricular rupture.
Recent studies show that effective (i.e., physiological) growth and remodeling of the heart's muscular tissue can reverse and cure CHF in some patients. Such a cure of CHF is called “ventricular recovery” herein, yet in some literature it is called reverse remodeling. The mechanism of ventricular recovery is a current and active area of research for the treatment of CHF. Hormonal and electro-physiological factors may play an important role. In addition, the influence of ventricular loading conditions on CHF is a significant factor. Unloading the diseased heart may decrease end-systolic volume and may create a more physiological strain pattern of the heart. A physiological strain pattern during contraction (also called “systole”) may lead to ventricular recovery, and it may prevent infarct expansion, myocardial scarring, and/or ventricular rupture.
Further evidence for a potential, fundamental role of strain in cardiac physiology and pathophysiology is that myocytes are highly sensitive to strain and respond with altered gene expression. (See Komuro and Yazaki, Ann. Rev. Physiol. 55:55-75(1993) and Sadoshima and Izumo, Ann. Rev. Physiol. 59:551-571(1997).) Numerous investigations have shown that altered hemodynamic loading and/or heart disease lead to growth and remodeling of myocytes and their extra-cellular matrix. It is also known that failing hearts exhibit a patho-physiological strain pattern where hoop strain and apex-base stain are more equal than normal and heart wall thickening during contraction is globally less than normal and more uniform across the wall. Although there is still debate on what is the normal motion or strain pattern during contraction, current results show a strain pattern wherein hoop strain is greater than apex-base strain and greater wall thickening is seen on the inner wall of the heart as compared to the outer wall.
Thus the present invention includes a method of restoring a more normal physiological strain pattern to a dyskinetic or failing heart, i.e. a heart with a patho-physiological strain pattern. By doing so, myocytes may be induced to grow in a normal manner. The effects on stress and strain on myocyte growth is described in greater detail in J. H. Omens, “Stress and strain as regulators of myocardial growth”, Prog. Biophys. Mol. Biol., 69(2-3:559-72 (1998).
Despite the promise shown by growth and remodeling of damaged or diseased heart tissue, there have been no known attempts to proactively modulate the strain pattern during contraction. Instead, treatments have been focused on ways to increase blood flow, off-load the heart, and/or reduce wall stress. These methods include some blood contacting assist devices, surgical reconstruction of the ventrilcle(s), cardiomyoplasty, and surgical insertion of passive devices. These treatments may promote a more healthy type of strain pattern during contraction, yet they do so indirectly if at all.
Current devices that provide direct mechanical assistance to the heart itself are often called direct cardiac compression devices (DCCDs), and they do modulate directly the kinematics during contraction. Yet current DCCDs do not proactively modulate the strain pattern so as to guide heart recovery and/or myocardial recovery. In contrast, many devices may cause detrimental remodeling and/or apoptosis because the induced strain pattern is so grossly abnormal. As is clear from the various patents and papers on these devices, current DCCDs have been optimized to promote systolic ejection, to be implanted easily, to reduce thrombo-embolic complication, to closely fit the heart contours during diastole, and criteria other than systolic strain pattern modulation.
An extremely important aspect of contraction strain pattern is the fact that it depends on both the end-diastolic configuration (reference configuration) AND the end-systolic configuration (current configuration). The strain field is a function of the gradient (with respect to reference position) of the mapping of material points from the reference configuration to the current configuration. Thus, the fact that prior DCCDs have tried to fit well the diastolic configuration is inconsequential to achieving an appropriate contraction strain pattern because their end-systolic configurations are either grossly aberrant or unknown.
First, an early DCCD (called a cardiac massager because it was designed for use in open chest surgery and not for implantation) developed by Vineberg is described in U.S. Pat. No. 2,826,193 (the Vineberg patent). This device, when inflated, produces a systolic state with low (or even inverted) curvature in the circumferential-radial plane of the heart (FIG. 2, Vineberg patent) and in the longitudinal-radial plane (FIG. 3, Vineberg patent). This is caused by the two opposite chambers which, when inflated, induce hourglass-like systolic geometries (FIG. 5, Vineberg patent). Thus the vineberg device induces an aberrant strain pattern during contraction.
A later device, the Anstadt cup, is described in U.S. Pat. No. 5,119,804 (the Anstadt patent). This device induces an inverted curvature of the longitudinal-radial plane during systole, but does retain a normal circumferential-radial plane curvature during systole (FIGS. 8, 9, and 10 Anstadt patent). The CardioSupport System made by Cardio Technolgies, Inc. is similar to the Anstadt cup and also induces curvature inversion in the longitudinal-radial plane.
Whereas the Anstadt cup inverts curvature in the longitudinal-radial plane, the heart booster (U.S. Pat. No. 5,713,954 and Ann. Thorac. Surg. 68:764-7) inverts curvature in the circumferential-radial plane, but not in the longitudinal-radial plane (FIGS. 1 and 2, Ann. Thorac. Surg. 68:764-7). This device prescribes a gear-like geometry with numerous spurs. Each spur in this end-systolic geometry induces curvature inversion during systole.
The Anstadt cup and heart booster attach to the apex of the heart, while the Vineberg device attaches to the base through a draw string constrictor at the valve plane. However, many other DCCDs attach to the interventricular grooves and either pull on the grooves (for example, the AbioBooster by Abiomed, Inc., Danvers, Mass. as described by Karvana et al., 2001, and the DCC Patch by Heart Assist Technologies, New South Wales, Australia) or hold them static (for example the device in U.S. Pat. No. 4,536,893 of Parravicini, the “Parravicini patent”). In either case, pulling or holding the grooves static likely decreases the curvature in the circumferential-radial plane during systole whereas for normal hearts curvature increases as the diameter decreases during systole. (Note that the radius-of-curvature R is the inverse of curvature C, i.e. C=1/R). Nevertheless, it is uncertain what the systolic configuration is for these devices because it is not disclosed.
The Abiobooster and DCC Patch cover one ventricle of the heart, and it is likely that when pressurized they decrease the curvature of the free wall and septum. The Parravicini device is sutured to the interventircular groove (or sulcus) and either pulls on the grooves using the two opposing membranes or is held static by the outer rigid shell. Again, the exact mode of operation is not clear from the Parravicini patent because a systolic configuration is not shown. In either case, the resulting strain pattern is not proactively modulated and a decrease in curvature likely occurs during contraction.
One embodiment (shown in FIG. 3, but not FIGS. 1 and 2) of the Hewson device shown in U.S. Pat. No. 3,034,501 (the Hewson patent) is similar to the Vineberg and Parravicini devices in that two opposing membranes squash the heart and tend to flatten it between the opposing membranes (i.e. induce a systolic configuration with decreased curvature in much of the circumferential-radial plane yet with dramatically increased curvature on the edges where the two membranes attach to the device). However, a systolic configuration is not shown.
Similarly, a systolic configuration is not shown for the embodiment of the Hewson device in FIGS. 1 and 2 of the Hewson patent. To derive a systolic configuration, one must consider a force balance with the pneumatic chamber pressurized. Upon doing a force balance, however, one realizes that the Hewson embodiment 1 is unrealistic and not implantable. A pressure P in the pneumatic chamber will induce at least an upward force of PA where A is the cross-sectional area of the opening rim. If there are tensile membrane stresses the upward force will be higher. Hewson suggests that contact forces between the rim and the heart are supposed to hold the heart in place, yet even active hearts are soft tissues that can undergo high shear and slip out of the Hewson device. Moreover, it is now known that myocardium is organized into sheets that allow contracting myocytes to rearrange and shear to attain high radial strains—i.e., motions that would allow the heart to slip out of the Hewson device. A hypocontractile or failed heart would be even more likely to slip out because it would more easily permit radial stain.
To evaluate this problem, a device similar to that of Hewson was constructed and implanted in a young bovine. Ultrasound measurements of the actual animal were used to precisely size the device. Even with normal contractility (i.e. high stiffness), the heart readily escaped from the device. The valve plane had to be sewn to the device and held in place with a stent that went from the device through the transverse pericardial sinus and pushed down on the commisure of the aortic and bicuspid valves.
Although a number of DCCDs are described above, other examples may be known to one skilled in the art. However, all current operable DCCDs suffer from a tendency to increase or even invert the curvature of the heart and thus produce an aberrant strain pattern during contraction.